The Next Small Thing

The first principle involved software. Hawkins, like so many of his rivals, had been trying to write code that was smart enough to enable PDAs to recognize all kinds of different handwriting styles. That meant the software had to be complex - which meant the PDA's performance had to be slow. (Nor, truth be told, could it be very good. Even the smartest handwriting software isn't very smart.) So why not reverse the logic? Why not ask people to learn some techniques to help the software better understand their handwriting?

"People are smarter than appliances," Hawkins told his colleagues. "They can learn.'' He recalled a lesson from his Berkeley days: "People like learning. People can learn to work with tools. Computers are tools. People like to learn how to use things that work."

Hence Graffiti, the handwriting-recognition software that differentiated PalmPilot from every other device on the market. In Graffiti, each letter is made by a single stroke of a pen on a small screen. Letters that normally require someone to lift a pen undergo slight changes: An A is written as an inverted V, an F as an inverted L. Graffiti requires just a small adjustment by the people who use it - but leads to a vast improvement in both accuracy and speed.

At home and in private, with pen and paper, Hawkins polished his letters, inscribing one atop the other to see how fast and how easily he could write them on a small surface area. When he unveiled Graffiti inside Palm, the reaction ranged from polite smiles to open opposition. "People were saying, 'Sure, we're going to ask people to learn a whole new way to write,' '' Colligan recalls. So Hawkins persuaded his colleagues to try learning the strokes. It took less time than they expected it to. "My sister is a technophobe," Colligan remembers. "She played around with it and said, 'Wow! It's cool.' ''

So much for the software. What about the device itself? Here Hawkins posed a simple question: How small is small enough? His answer yielded the second principle: Small enough to fit in a shirt pocket. He paced the hallways at Palm headquarters, ruler in hand, measuring pocket sizes against small blocks of balsa wood. He designed screens and pasted down configurations of various applications. He pushed to simplify: Which features are mandatory? Which can be sacrificed? Which might be optional? With each revision, the product kept getting smaller.

By August 1994 - less than three months after Hawkins began rethinking the market - Palm Computing had a mockup of its new device. The product would fit in a shirt pocket. It would run on AAA batteries. It would offer four core functions: a calendar, an address book, a To-Do list generator, and a memo-writing feature. It would sell for less than $300. It was a marvel of elegance and simplicity.

"I remember sitting around the table when it was presented," says Colligan. "We all had goose bumps."

The product got a code name: Touchdown. Palm Computing got a new company mantra: "No Excuses.''

New Product, New Company

It's one thing to learn from mistakes. it's quite another to persuade investors that you can get things right the second time. "We had been working on this product for two years," says Hawkins. "Donna, Ed, and I really believed in what we had come up with. Touchdown was it. But the market kept getting worse and worse."

The time had come for another course correction. It began in a conference room next to Bruce Dunlevie's office at Merrill Pickard. Dubinsky and Hawkins arrived dejected. They told Dunlevie that they finally had the right product, but that their previous partners wouldn't get behind it. Dunlevie - soft-spoken, circumspect, bookish - listened, gauged their despair, and said, "Stop complaining. You know how to do this, right?''

Hawkins said he thought so.

"Then do the whole product,'' Dunlevie told Hawkins. "Just go out and do it.''

It was radical advice. Dunlevie was arguing that the company couldn't redesign its product unless it reinvented its business model. Palm would no longer be just a software company. It would be a self-sufficient company that designed, built, and marketed Touchdown. "I didn't think that doing the whole product - both the hardware and the software - was an option," Hawkins says. "It was expensive. I was a young guy. I had never thought about doing anything that big alone."

Which didn't mean that Palm would build factories and warehouses, or hire hundreds of salespeople to call on retailers. (Indeed, as amazing as it sounds, the company hired just one new employee in 1994.) Instead, Hawkins and Dubinsky assembled a dispersed team of hardware-design and contract-manufacturing companies, each of which brought resources to the table. They paid the companies with modest sums of cash, lots of stock options, and the promise of future glory. "We created a virtual company,'' Hawkins says.